PreprintPDF Available

Direct Pathway Neurons in Mouse Dorsolateral Striatum In Vivo Receive Stronger Synaptic Input than Indirect Pathway Neurons

Authors:

Abstract

Striatal projection neurons, the medium spiny neurons (MSNs), play a crucial role in various motor and cognitive functions. MSNs express either D1 or D2 type dopamine receptors and initiate the direct-pathway (dMSNs) or indirect pathways (iMSNs) of the basal ganglia, respectively. dMSNs have been shown to receive more inhibition than iMSNs from intrastriatal sources. Based on these findings, computational modelling of the striatal network has predicted that under healthy conditions dMSNs should receive more excitatory input than iMSNs. To test this prediction, we analyzed in vivo whole-cell recordings from dMSNs and iMSNs in healthy and dopamine-depleted (6OHDA) anaesthetized mice. By comparing their membrane potential fluctuations, we found that dMSNs exhibited considerably larger membrane potential fluctuations over a wide frequency range. Furthermore, by comparing the spike-triggered average membrane potentials, we found that dMSNs depolarized towards the spike threshold significantly faster than iMSNs did. Together, these finding corroborate the theoretical prediction that direct-pathway MSNs receive stronger input than indirect-pathway neurons. Finally, we found that dopamine-depleted mice exhibited no difference between the membrane potential fluctuations of dMSNs and iMSNs. These data provide new insights into the question how a lack of dopamine may lead to behavior deficits associated with Parkinson’s disease. Significance statement The direct and indirect pathways of the basal ganglia originate from the D1 and D2 type dopamine receptor expressing medium spiny neurons (dMSNs and iMSNs), respectively. To understand the role of the striatum in brain function and dysfunction it is important to characterize the differences in synaptic inputs to the two MSN types. Theoretical results predicted that dMSNs should receive stronger excitatory input than iMSNs. Here, we studied membrane potential fluctuation statistics of MSNs recorded in vivo in anaesthetized mice and found that dMSNs, indeed, received stronger synaptic input than iMSNs. We corroborated this finding by spike-triggered membrane potential analysis, showing that dMSNs spiking required more synaptic input than iMSNs spiking did, as had been predicted by computational models.
Filipović et al. Comparison of total input to striatal MSNs in vivo
Direct Pathway Neurons in Mouse Dorsolateral Striatum In Vivo Receive Stronger
Synaptic Input than Indirect Pathway Neurons
Marko Filipović
1,2
, Maya Ketzef
3
, Ramon Reig
4
, Ad Aertsen
2
, Gilad Silberberg
3
, Arvind
Kumar1,2
These two authors contributed equally to this work.
1
Dept. of Computational Science and Technology, School of Computer Science and Commu-
nication, KTH Royal Institute of Technology, Stockholm, Sweden
2Bernstein Center Freiburg, University of Freiburg, Germany
3Dept. of Neuroscience, Karolinska Institute, Stockholm, Sweden
4
Instituto de Neurociencias, Consejo Superior de Investigaciones Científicas & Universidad
Miguel Hernández, San Juan de Alicante, Spain
Corresponding author: Arvind Kumar (arvkumar@kth.se)
Abstract
Striatal projection neurons, the medium spiny neurons (MSNs), play a crucial role in various
motor and cognitive functions. MSNs express either D1 or D2 type dopamine receptors
and initiate the direct-pathway (dMSNs) or indirect pathways (iMSNs) of the basal ganglia,
respectively. dMSNs have been shown to receive more inhibition than iMSNs from intrastriatal
sources. Based on these findings, computational modelling of the striatal network has predicted
that under healthy conditions dMSNs should receive more excitatory input than iMSNs. To
test this prediction, we analyzed in vivo whole-cell recordings from dMSNs and iMSNs in
healthy and dopamine-depleted (6OHDA) anaesthetized mice. By comparing their membrane
potential fluctuations, we found that dMSNs exhibited considerably larger membrane potential
fluctuations over a wide frequency range. Furthermore, by comparing the spike-triggered
average membrane potentials, we found that dMSNs depolarized towards the spike threshold
significantly faster than iMSNs did. Together, these finding corroborate the theoretical
prediction that direct-pathway MSNs receive stronger input than indirect-pathway neurons.
Finally, we found that dopamine-depleted mice exhibited no difference between the membrane
potential fluctuations of dMSNs and iMSNs. These data provide new insights into the question
how a lack of dopamine may lead to behavior deficits associated with Parkinson’s disease.
Significance statement
The direct and indirect pathways of the basal ganglia originate from the D1 and D2 type
dopamine receptor expressing medium spiny neurons (dMSNs and iMSNs), respectively. To
understand the role of the striatum in brain function and dysfunction it is important to
characterize the differences in synaptic inputs to the two MSN types. Theoretical results
predicted that dMSNs should receive stronger excitatory input than iMSNs. Here, we studied
1
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
membrane potential fluctuation statistics of MSNs recorded in vivo in anaesthetized mice and
found that dMSNs, indeed, received stronger synaptic input than iMSNs. We corroborated this
finding by spike-triggered membrane potential analysis, showing that dMSNs spiking required
more synaptic input than iMSNs spiking did, as had been predicted by computational models.
Introduction
The striatum is the largest nucleus in the basal ganglia (BG) and acts as its main input
structure. GABAergic medium spiny neurons (MSNs) are the striatal projection neurons
and constitute about 95% of the striatal neuronal population. D1 type dopamine receptor
expressing MSNs (dMSNs) project to the substantia nigra pars reticulata and globus pallidus
interna and constitute the ’direct pathway’, whereas D2 type dopamine receptor expressing
MSNs (iMSNs) project to the globus pallidus externa and constitute the ’indirect pathway’. A
balance in the activity of the two pathways is essential for correct functioning of the BG, and
is disrupted in BG-related pathologies such as Parkinson’s disease (PD). To understand how
the direct and indirect pathways shape BG function, we need to quantify both the upstream
excitatory inputs into the striatum and the recurrent inhibitory connections within and between
dMSNs and iMSNs.
The dMSNs and iMSNs differ in their connectivity: iMSN to dMSN connectivity (13%) is
much higher than dMSN to iMSN (4.5%), whereas dMSN to dMSN connectivity (7%) is
much lower than iMSN to iMSN (23%) (Taverna et al., 2008; Planert et al., 2010). Moreover,
GABAergic fast-spiking interneurons (FSIs) connect preferentially to dMSNs compared to
iMSNs (53% vs. 36%) (Gittis et al., 2010). That is, dMSNs receive overall more inhibition
than iMSNs. Despite these differences, both dMSNs and iMSNs exhibit similar average activity
in awake behaving animals (Cui et al., 2013; Sippy et al., 2015).
Using a computational model we recently predicted that dMSNs should receive stronger
excitatory input than iMSNs (either through more synapses, stronger synapses, or stronger
input rates and/or correlations), so that both dMSNs and iMSNs may have comparable firing
rates (Bahuguna et al., 2015). Recent ex vivo recordings suggest that cortico-striatal synapses
on dMSNs may be stronger than those on iMSNs (Parker et al., 2016) (however, see Lei et al.
(2004); Kress et al. (2013); Doig et al. (2010); Deng et al. (2015)). While this data supports
the theoretical predictions, it is well known that in vivo synaptic conductances can be very
different from ex vivo measurements (Destexhe et al., 2003).
Even though it is hard to estimate the full strength and numbers of individual excitatory
synapses impinging on dMSNs and iMSNs experimentally, a relative difference in the total
input to the two neuron types can be estimated by analyzing in vivo intracellular membrane
potential fluctuations. In particular, the variance (or the spectral power) of the membrane
potential fluctuations is proportional to the square of the synaptic strength (Kuhn et al., 2004).
That is, by comparing the spectra of sub-threshold membrane potential in vivo we can test
whether dMSNs indeed receive stronger excitatory input than iMSNs, as was theoretically
predicted (Bahuguna et al., 2015).
Therefore, we recorded and analyzed the in vivo membrane potentials of dMSNs and iMSNs from
healthy and dopamine-depleted anaesthetized mice using whole-cell patch clamp recordings.
These neurons exhibited alternating periods of high and low activity (called up- and down-
states, respectively), characteristic of recordings in animals under ketamine-induced anaesthesia
2
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
(Wilson and Kawaguchi, 1996). We found that dMSNs exhibited higher spectral power in their
up-states than iMSNs over a wide range of frequencies in healthy mice. In addition, bilateral
whisker stimulation in healthy animals showed that sensory inputs evoked larger responses
in dMSNs than in iMSNs. Despite these differences, the membrane time constants of the
two MSN types were not significantly different. According to linear systems theory, stronger
membrane potential fluctuations are indicative of stronger synaptic inputs and/or higher input
correlations (Kuhn et al., 2004). Finally, we found that dopamine depletion abolished the
difference in spectral power of up-state membrane potential fluctuations between dMSNs and
iMSNs, highlighting the role of dopamine in maintaining the activity balance between the
direct and indirect pathways.
Thus, our study provides the first experimental in vivo evidence of stronger synaptic input to
the direct-pathway of the mouse dorsolateral striatum, and demonstrates that this difference
is attenuated in dopamine-depleted animals.
Methods
Experimental Methods
Ethics approval. All experiments were performed according to the guidelines of the Stockholm
municipal committee for animal experiments under an ethical permit to G.S. (N12/15).
D1-Cre (EY262 line) or D2-Cre (ER44 line, GENSAT) mouse line were crossed with the
Channelrhodopsin (ChR2)-YFP reporter mouse line (Ai32, Jackson laboratory) to induce
expression of ChR2 in either dMSNs or iMSNs, respectively. Mice of both sexes were housed
under a 12-hour light-dark cycle with food and water ad libitum. All experiments were carried
out during the light phase.
6OHDA lesioning. Mice (12 males and females 8-10 weeks of age) were anesthetized with
isoflurane and mounted in a stereotaxic frame (David Kopf Instruments, Tujunga, California).
The mice received one unilateral injection of
1µL
of 6OHDA-HCl (
3.75 µg/µL
dissolved in
0.02 %
ascorbic acid) into the medial forebrain bundle (MFB), according to the following
coordinates (Paxinos and Franklin, 2004): antero-posterior
1.2 mm
, medio-lateral
1.2 mm
and dorso-ventral
4.8 mm
. After surgery, all mice were injected with Temgesic (
0.1 mg/kg
,
Reckitt Benckiser, Berkshire, England) and allowed to recover for at least 2 weeks. Sham and
unlesioned mice (n = 21 of both sexes) served as controls, their data were pooled after no
differences were found between the groups. Only 6OHDA injected mice that showed rotational
behavior (Santini et al., 2007) were used in our experiments.
In vivo recordings. Experiments were conducted as described previously (Reig and Silberberg,
2014; Ketzef et al., 2017). Briefly, 2-3 weeks post-lesioning, mice were anesthetized by
intraperitoneal (IP) injection of ketamine (
75 mg/kg
) and medetomidine (
1 mg/kg
) diluted
in
0.9%
NaCl. To maintain mice under anesthesia, a third of the dose of ketamine was
injected intraperitonally approximately every 2 hours or in case the mouse showed response to
pinching or changes in EcoG patterns. Mice were tracheotomized, placed in a stereotactic
frame, and received oxygen enriched air throughout the recording session. Core temperature
was monitored with a feedback-controlled heating pad (FHC) and kept on
36.5±0.5C
. Patch
clamp recordings were performed in the dorsolateral striatum since the sensory and motor
areas project topographically onto it (McGeorge and Faull, 1989). The skull was exposed
3
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
and a craniotomy was drilled (Osada success 40)
3.5
-
4 mm
lateral to the bregma, and the
dura was removed. Patch pipettes were pulled with a Flaming/Brown micropipette puller
P-1000 (Sutter Instruments). Pipettes (
7
-
10 M
, borosilicate, Hilgenberg), back-filled with
intracellular solution, were inserted with a ~
1500 mbar
positive pressure to a depth of about
2 mm
from the surface, after which the pressure was reduced to
30
-
35 mbar
. The pipette
was advanced in
1µm
steps in depth (35 degrees angle), in voltage clamp mode. When a
cell was encountered, the pressure was removed to form a Gigaseal, followed by application
of a ramp of increasing negative pressure until a cell opening was evident. Recording was
performed in current clamp mode. Intracellular solution contained: 130 K-gluconate, 5 KCl,
10 HEPES, 4 Mg-ATP, 0.3 GTP, 10 Na2-phosphocreatine, and
0.2
-
0.3%
biocytin (pH = 7.25,
osmolarity 285mOsm). The exposed brain was continuously covered by
0.9%
NaCl to prevent
drying. Signals were amplified using a MultiClamp 700B amplifier (Molecular Devices) and
digitized at
20 kHz
with a CED acquisition board and Spike 2 software (Cambridge Electronic
Design).
Optogenetic identification of in vivo recorded neurons. To obtain “on line” identification of
whole-cell recorded neurons, we used the optopatcher (Katz et al., 2013) (A-M systems, WA
USA). Computer controlled pulses of blue light (
7 mW
LED,
470 nm
, Mightex systems) were
delivered through an optic fiber inserted into the patch-pipette while recording the responses
in whole-cell configuration (Fig. 1A). Light steps (
500 ms
) were delivered every 2-5 seconds
with increasing intensity between 20 to
100 %
of full LED power (
2.1 mW
at the tip of the
fiber). Positive cells responded to light stimulation by step-like depolarization with or without
firing, whereas negative cells did not show any response (Fig. 1B, and see Ketzef et al. (2017)
for full characterization).
Whisker stimulation. Air puffs were delivered by a picospritzer (Picospritzer III, Parker Hannifin)
through plastic tubes (
1 mm
diameter) positioned up to a centimeter from the mouse’s whiskers.
Air puff stimulations (
15 ms
) were delivered at
0.2 Hz
and at least 30 responses were acquired
for each stimulation condition. The air pressure was set to 103.4-137.9 kPa (15-20 PSI).
Data Analysis
Up- and down-state detection. For each membrane potential recording, we used a short time
window (
20
-
100 ms
, depending on the noise level in the recording) to identify sudden transitions
in the membrane potential with an amplitude sufficiently large to cross the cell-specific up-state
or down-state thresholds. Upon detection of such a transition, we classified the following
voltage period as an up-state or a down-state (Fig. 2A). The next sufficiently large membrane
potential transition in the opposite direction marked the ending of that state. State thresholds
were determined by finding the two main peaks of the bimodal voltage histogram of the
entire trace, and by empirically adjusting these thresholds for the best detection rate (see also
Léger et al. (2005) and Fucke et al. (2011)). In cells where the overall baseline voltage level
fluctuated over time, we either used only the most stable section of the recording or discarded
the entire recording alltogether. All states with a duration shorter than
40 ms
were discarded
from the analysis (Mukovski et al., 2007).
For the purpose of characterizing the sub-threshold membrane potential dynamics analysis
we excluded all up-states during which spiking occurred. Moreover, we also excluded a state
from further analysis if one or more of the following criteria was met: (1) an up-state was
4
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
Fig 1. MSNs classification using the optopatcher. To facilitate the classification of MSNs as belonging to either the direct
(dMSN) or indirect (iMSN) pathway in vivo, we utilized an optogenetic approach. In either D1-Cre or D2-Cre animals crossed
with ChR2 reporter mouse, we selectively expressed ChR2 in dMSNs or iMSNs, respectively. Using the optopatcher, we could
deliver focal light stimulation to the recorded cell and classify its identity ’online’ during whole cell patch recordings. A
Illustration of the experimental approach (left). In anesthetized mice, the optopatcher is introduced through the craniotomy. The
optic fiber is inserted into the patch pipette and light application is focal. MSNs of both pathways are intermingled (right),
positive cells (green) express ChR2 and YFP, whereas negative cells (black) do not.
B
Whole cell patch recording from positive
(left) and negative (right ) cells in a D2-ChR2 mouse. When the blue light is activated (470 nm,0.5 s), positive cells depolarize
immediately, whereas negative cells are not affected. Each example shows 10 repetitions (gray), overlaid by the average trace
(green for positive and black for negative cells).
interrupted by a down-state shorter than
25 ms
(both states were discarded), (2) the mean
membrane potential of a down-state exceeded the global average down-state potential for
that cell by more than 3%, (3) a recording artefact was present, or (4) a whisker stimulation
trigger occurred either during the state or 200 ms preceding the state.
Finally, for all remaining states, we removed
5%
of the data, from the start of the state and
before the end of the state, to minimize the impact of state transitions on the measured
variables.
Power spectral density (PSD) estimation. The PSD estimate of an up-state membrane voltage
trace was determined by first subtracting the mean potential from the remainder of the trace,
and then by applying MATLAB’s periodogram function with Bartlett-Hann windowing. The
minimal detectable frequency in individual up-state PSDs was set as the inverse of the duration
of that state. For each cell, all such up-state PSD estimates were averaged to obtain a single
power spectral density curve (Fig. 2C, gray traces). When comparing PSDs across cell groups
(dMSNs vs. iMSNs), we constructed a grand-average PSD for each group by averaging over
PSDs of individual cells (Fig. 2C, color traces). Frequencies below
5 Hz
were disregarded
because we observed only few up-states longer than
200 ms
. Additionally, all frequency content
between
45
and
55 Hz
was removed to avoid power line contamination. We restricted the
higher frequency range to
150 Hz
, adopting this as the upper limit of the high-gamma band in
our study.
We divided PSD estimates obtained in this manner into five frequency bands: sub-
α
(
5
-
8 Hz
),
α
(
8
-
13 Hz
),
β
(
13
-
30 Hz
), low-
γ
(
30
-
70 Hz
), and high-
γ
(
70
-
150 Hz
). To calculate the total
5
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
power within any one frequency band for each cell, we isolated the section of interest of the
PSD estimate and integrated the area under the curve.
Due to very low levels of spectral power during the down-states, the line noise power precluded
any meaningful PSD comparison across the two cell groups.
Effective membrane time constant estimation. We estimated the effective membrane time
constant
τm
from the in vivo membrane potential fluctuations by the following method.
For a narrow enough voltage range, the membrane of a neuron can be approximated as a
linear low-pass filter. Then, for that narrow voltage range,
τm
is directly proportional to the
membrane capacitance and inversely related to the total membrane conductance. Thus, to
minimize non-linear voltage-dependent effects, to account for the voltage dependence of
τm
(Kuhn et al., 2004), and to be able to treat the neuron membrane as a linear low-pass filter,
we first binned the average membrane potentials of all states in
0.5 mV
wide bins. Then we
estimated the power spectral density of individual states belonging to each bin and averaged
over the estimates in order to reduce noise, as explained in the previous section. Further noise
reduction was achieved by smoothing the averaged PSD estimate with a Gaussian kernel, and
the resulting curve was used to extract the cutoff frequency
fc
, calculated as the point where
the maximal value of the smoothed PSD estimate fell to one half (
3 dB
point, Fig. 3A). The
initial effective membrane time constant
τini
m
was then calculated as 1
/
(2
πfc
). We repeated
this procedure for a series of narrow voltage ranges across different instances of up- and
down-states within a single cell, in order to avoid non-linearities induced by large excursions of
the effective membrane conductance.
The smoothing of the average PSD estimate introduces a shift of the
3 dB
point, leading
to an erroneous estimation of the effective membrane time constant. The magnitude and
sign of the error depend, in a non-linear fashion, on the width of the Gaussian kernel used
for smoothing in the frequency domain, and the duration of the original signal in the time
domain. To account for this error, we numerically determined a correction term
τC
m
, which we
could then add to the initially estimated value τini
m, to obtain the final MSN membrane time
constant estimate
τm
. This correction term was calculated as follows. We constructed multiple
surrogate “neuronal” time series by filtering Gaussian white noise signals of different durations
through a set of low-pass Butterworth filters (third order, zero-phase) with predetermined
cutoff frequencies. Thus, for each of the surrogate time series we knew the actual time
constant (
τactual
) of the underlying low-pass filter. We then proceeded to make an initial
estimate of the time constant (
τini
) as described above, using a single fixed value for the
kernel width of the Gaussian smoothing function (
kw
= 12). The error term was then defined
as
τC
=
τactual τini
. Using this approach, we obtained the correction term
τC
for signals of
different durations and filters with different time constants. Next, we defined
τC
as a function
of
τini
and signal duration (Fig. 3B) to obtain the correction term
τC
m
for our estimates of the
MSN membrane time constant. Finally, the effective MSN membrane time constant
τm
was
determined by adding the corresponding τC
mto the initial estimate τini
m.
The main weakness of this method stems from the necessity of averaging the spectral data
over many trials of sufficient duration and power. That is, for the most precise estimation, the
trials (states) should preferably be at least
250 ms
long, and the input to the neurons should
have rich enough frequency content to uncover the membrane cutoff frequency (comparable
to injecting white noise into the in vitro recorded neuron).
Due to the underlying approximations and limitations of the method, the estimated values of
6
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
the membrane time constants should not be treated as actual, precise values of those neurons’
τm
. Nevertheless, our approach does return consistent and comparable results across different
cells when applied to the recorded data. Moreover, our analysis employing the
τm
estimation
procedure uncovers differences in membrane time constant of the down-states similar to those
previously reported in Ketzef et al. (2017) (Fig. 3D).
Spike-triggered average (STA) calculations. For every recorded neuron that spiked we extracted
12 ms
of the pre-spike voltage traces. The duration of this particular time window was chosen
as it roughly represents the average membrane integration window for synaptic input in the
up-states, based on the estimations of the effective membrane time constants across different
cell groups (Fig. 3D). For every cell, spikes were identified in the voltage trace, and the intervals
from
0.25 ms
before to
5 ms
after the spike events were removed from the trace. The spiking
threshold was then determined as the largest fluctuation of the first derivative of the remaining
trace. The times when the derivative of the full trace crossed the threshold were taken as
spike onset times. For the purpose of calculating the average of these pre-spike voltage traces
(STA), we did not include any spikes occurring during state transitions, that initiated less
than
12 ms
after the start of an up-state, or which were occurred earlier than
17 ms
after the
previous spike in the same up-state. The remaining pool of spikes was divided into those that
were the result of spontaneous neuronal activity and those that arose as the consequence of
whisker stimulation. If, after these selections. the pool contained at least three spikes, the
STA was calculated.
The STAs were compared using a permutation test. For each group comparison we collected
all the cell-average traces into a single pool, shuffled their indices, and generated randomized
groups by drawing as many traces from this common shuffled pool as the original groups
had. For each such generated randomized group, we constructed a grand-average STA. We
repeated this process 1,000 times. Significance lines were determined as the
2.5%
and
97.5%
of the voltage distributions of the random grand-average traces for each time point. The
range of voltage distributions differed between groups when the number of traces belonging
to randomized groups for a single comparison was different (e.g., for the comparison of
spontaneous vs. evoked iMSN STAs, we had 11 spontaneous and 4 evoked cell-average traces).
This difference is reflected in the voltage ranges depicted in the graphs, but it does not affect
the validity of the permutation test.
Statistical methods. nless noted otherwise, the data are presented as mean
±
SEM and were
tested for normality using the Shapiro-Wilk test. Normally distributed data were tested by
the unpaired two-sample Student’s t-test, and non-normally distributed data by the Wilcoxon
rank-sum test (ttest2 and ranksum in MATLAB, respectively). The significance level
α
was
set to 0.05. In the case of PSD comparison over different frequency bands (Fig. 2D), the
results were corrected for multiple testing by the Holm-Bonferroni correction (Holm, 1979),
and both the corrected α-level (αHB ) and the calculated p-value are reported.
All data analyses were performed using custom scripts written in MATLAB R2016a (Mathworks,
Inc.).
Results
To estimate the relative strength of excitatory synaptic inputs to striatal neurons, we obtained
in vivo whole-cell patch clamp recordings of MSNs from the dorsolateral striatum in control
7
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
(dMSN
n
= 26, iMSN
n
= 18, total
n
= 44) and 6OHDA lesioned mice (dMSN
n
= 14,
iMSN
n
= 9, total
n
= 23). We used optogenetic stimulation to classify MSNs online
during the recording session as belonging to either the direct or indirect pathway using the
optopatcher (Katz et al., 2013). Both dMSNs and iMSNs showed slow-wave membrane
potential oscillations (up- and down-states), characteristic of neurons recorded in animals
under ketamine-induced anesthesia (Wilson and Kawaguchi, 1996) (Fig. 2A). During the
up-state, MSNs receive barrages of excitatory inputs from the neocortex and the thalamus.
We therefore analyzed the variance and spectrum of the up-state membrane potential traces
to assess the respective synaptic inputs to the two MSN types.
dMSNs have higher spectral power in up-state than iMSNs
If a neuron soma is treated as a simple linear integrator, the mean and variance of the
subthreshold membrane potential fluctuations is primarily determined by the firing rate, the
number of excitatory and inhibitory inputs to a given cell, and their synaptic strength (Kuhn
et al., 2004):
µv=Ur+λeZEPSP(t)λiZIPSP(t)dt
σ2
v=λeZEPSP(t)2dt +λiZIPSP(t)2dt (1)
where
µv
and
σ2
v
are the voltage-dependent mean and variance of the membrane potential,
Ur
is the resting membrane potential,
λe
and
λi
are the rates of excitatory and inhibitory
inputs, respectively, and
EPSP
(
t
)and
IPSP
(
t
)describe the temporal shape of excitatory and
inhibitory post-synaptic potentials.
From Eq. 1 it is clear that excitatory and inhibitory inputs have an opposite effect on the
mean of the membrane potential of a cell receiving synaptic input. By contrast, because
the calculation of the variance involves the square of the PSPs kernel, an increase in either
excitatory or inhibitory inputs always results in an increase of the variance of the membrane
potential (Eq. 1). Against this background, consider two neurons,
ns
and
nw
, receiving inputs
via stronger and weaker synapses, respectively. The excitatory and inhibitory inputs to these
two neurons can be tuned such that both
ns
and
nw
have the same mean membrane potential.
However, due to the stronger synaptic weights and, hence, larger post-synaptic potentials, the
neuron
ns
will exhibit a larger membrane potential variance than the neuron
nw
. This example
illustrates that the mean membrane potential is not an adequate measure for the overall
synaptic input, but by comparing the variances it is possible to determine if two neurons receive
different amounts of synaptic inputs. This requires that the two neurons receive uncorrelated
synaptic inputs and that their membrane time constants are similar.
Since the variance in time-domain equals the power spectral density (PSD) in frequency
domain (Parseval’s theorem), the PSD gives an estimate of the variance for every frequency in
the signal (Papoulis and Pillai, 2002). Therefore, we measured the PSD of the membrane
potential for every detected up- and down-state of a cell (Fig. 2A, see Methods).
For each MSN type we constructed a grand-average PSD estimate for both control and
6OHDA conditions (Fig. 2B). Direct comparison of these grand-averages revealed that dMSNs
had consistently higher PSD than iMSNs over all examined frequency bands under control
conditions. In particular, in three prominent, higher-frequency bands (
β
:
13
-
30 Hz
,
Z
= 2
.
47,
8
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
p
= 0
.
0135,
αHB
= 0
.
0167; low-
γ
:
30
-
70 Hz
,
t(42)
= 2
.
72,
p
= 0
.
0095,
αHB
= 0
.
01;
and high-
γ
:
70
-
150 Hz
,
Z
= 2
.
57,
p
= 0
.
01,
αHB
= 0
.
0125) dMSNs showed significantly
higher power than iMSNs (Fig. 2C, top left). Because the total power spectral density of
the membrane potential in a selected frequency band equals the variance of the membrane
potential in that frequency band (Papoulis and Pillai, 2002), the heightened power of dMSN
up-state membrane potentials in control animals is indicative of stronger voltage fluctuations
as compared to iMSNs. Unlike under control conditions, in the DA-depleted striatum we found
no difference in the spectral power of up-state membrane potential fluctuations of dMSNs
and iMSNs (across all bands
p >
0
.
68, Fig. 2C, top right). Comparison of the up-state control
vs. DA-depleted conditions revealed a significant difference in dMSN
α
-band (
8
-
13 Hz
) power
(
Z
= 2
.
62,
p
= 0
.
0087,
αHB
= 0
.
01). DA depletion did not affect the spectral power of
iMSNs (control vs. DA-depleted conditions across all bands
p >
0
.
55). Finally, in down-states,
there was no significant difference in spectral power of dMSNs and iMSNs in either condition
(the calculated significance value was always above the corrected alpha level).
Given that up-states are thought to be primarily synaptically driven (Wilson and Kawaguchi,
1996; Stern et al., 1997), our results indicate that the increased power of dMSNs, especially
in the higher-frequency bands, compared to iMSNs in the control case stems from stronger
total input to direct pathway striatal neurons. Furthermore, our results suggest that in
dopamine-depleted conditions, the total input to dMSNs is either significantly reduced and/or
is more similar to the input to iMSNs.
MSN membrane time constant does not underlie the differences in high-frequency
power
The difference in high-frequency power between dMSNs and iMSNs may be caused by a
difference in the time constants of the two neuron types. We estimated the effective time
constant using the spectrum of the membrane potential fluctuations (see Methods).
We found that the effective time constants for both dMSNs and iMSNs in the up-states were
smaller than in the down-states (Table 1, Fig. 3C). On average, the ratio of down-state to
up-state effective membrane time constant across all groups was 1
.
86 (Fig. 3D, dMSN control
1
.
76, iMSN control 1
.
85, dMSN 6OHDA 2
.
05, iMSN 6OHDA 1
.
77). This is similar to the
case of neocortical neurons, which also show a shorter time constant in up-states (Paré et
al., 1998; Destexhe et al., 1999; Léger et al., 2005). However, in MSNs this ratio is not as
large as has been reported for neocortical neurons (Reig and Silberberg, 2014), presumably
because of the closing of potassium inward rectifier (Kir) channels in MSNs, happening as the
membrane depolarizes (Waters and Helmchen, 2006; Nisenbaum and Wilson, 1995).
Further comparisons showed no significant difference between the up-state effective time
constants of dMSNs and iMSNs in control or 6OHDA conditions (in all cases
p >
0
.
085;
Fig. 3D). However, in the down-states, the effective
τm
of dMSNs was slightly larger in the
6OHDA condition than in the control (
t(38)
=
2
.
05,
p
= 0
.
048; control
n
= 26, 6OHDA
n
= 14), whereas no such difference was present for the iMSNs (
p
= 0
.
37; Fig. 3D). These
results are partially consistent with previously reported measurements of input resistance using
standard methods in MSN down-states (Ketzef et al., 2017).
Taken together, these results clearly suggest that the differences in the power spectra of
up-state sub-threshold membrane potential fluctuations between dMSNs and iMSNs (Fig. 2D)
9
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
Fig 2. dMSNs carry more power than iMSNs during up-states in control conditions. ALeft : Ten seconds of membrane
potential recordings for a dMSN in control (upper trace) and 6OHDA (lower trace) conditions, exhibiting up- and down-states
(green and red, respectively). Dashed lines represent the two cell-specific voltage thresholds used for state classification (see
Methods). Right: Distributions of membrane potential values for the entire recordings of the two neurons shown at left. Note the
characteristic bimodality of the up- and down-states.
B
Grand-average PSD estimates of up-states for all dMSNs and iMSNs in
control (top, red and blue, respectively) and 6OHDA (bottom, light red and light blue, respectively) conditions. Grey traces
represent average up-state PSD estimates of individual neurons. Frequencies between
45
and
55 Hz
were removed to avoid power
line contamination (see Methods).
C
Comparison of grand-average PSD estimates in different frequency bands. dMSNs exhibited
higher power spectral density than iMSNs in control conditions in beta (p= 0.0135,αHB = 0.0167), low-gamma (p= 0.0095,
αHB
= 0
.
01), and high-gamma bands (
p
= 0
.
0103,
αHB
= 0
.
0125; dMSN
n
= 26, iMSN
n
= 18 for all three bands), indicating
either stronger or more frequent synaptic input. dMSNs also showed increased PSD in control versus 6OHDA for the 8-13 Hz
band (p= 0.0087,αHB = 0.01). Test statistics were corrected using the Holm-Bonferroni procedure.
are not the result of different membrane time constants of the two types of neurons. Moreover,
the lower membrane time constant of MSNs in the up-state suggests that these neurons also
operate in a relatively high conductance regime.
10
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
Table 1.
Comparison of the effective time constants of dMSNs and iMSns in the up-states vs. down-states. For both MSN types,
across both healthy and dopamine-depleated conditions, up-states exhibited significantly faster membrane dynamics than
down-states.
τup
m(ms) τdown
m(ms) nr. samples statistics p-value
control dMSN 12.7±0.3 22.4±1.026 -6.0 (Z)1.8×109
iMSN 12.3±0.3 22.8±0.718 -13.1 (t34)7.1×1015
6OHDA dMSN 12.8±0.7 26.2±1.714 -7.3 (t26)9.5×108
iMSN 13.7±0.8 24.3±1.79 -5.5 (t16)4.8×105
dMSNs receive stronger input from mouse sensory cortex than iMSNs
In the analysis so far we focused on subthreshold membrane potential sections, in which
neurons did not fire action potentials during the up-states. To further test the hypothesis that
dMSNs indeed receive stronger inputs than iMSNs, we investigated the membrane fluctuations
leading to action potential discharges. To this end, we obtained the spike-triggered average
(STA) of the membrane potential immediately preceding action potential discharge for each
neuron (Fig. 3A-C). If dMSNs would indeed receive stronger inputs, we would expect the
corresponding STA traces to approach the spike threshold with a steeper slope, compared to
the STA traces of iMSNs. To better quantify this difference, we sub-divided the spikes of each
neuron into those corresponding to spontaneous spiking activity and those evoked by whisker
deflections with brief air puffs. This STA analysis was only performed for healthy animals, as
in our data MSNs recorded from DA-depleted mice elicited only a very small number of spikes,
not sufficient for STA analysis.
Comparison of the grand-average STAs for the two MSN types upon bilateral whisker stimulation
revealed that dMSNs indeed depolarized to the spike threshold much faster than iMSNs. For
the spontaneously generated spikes, this applied as well, although the difference was less
prominent. Comparing the average membrane potential
12 ms
before spike time (the duration
of the average up-state integration window shown in Fig.3D), we found that dMSN membrane
potentials were on average
1.3 mV
more hyperpolarized than those of iMSNs, resulting in
steeper depolarization slopes (
k
) preceding spike initiation (Fig.4D,
t
=
12 ms
; dMSN,
6.27±0.34 mV
,
k
=
0.31 mV/ms
,
n
= 10; iMSN,
4.97±0.53 mV
,
k
=
0.28 mV/ms
,
n
= 10). This difference was even bigger for whisker stimulation evoked spikes (Fig. 4E,
t
=
12 ms
; dMSN,
9.95±1.16 mV
,
k
=
0.58 mV/ms
,
n
= 5; iMSN,
6.01±0.72 mV
,
k=0.34 mV/ms,n= 4).
We further examined the STA differences between dMSNs and iMSNs by utilizing a permutation
test (see Methods). When a grand-average STA trace would fall above
97.5%
or below
2.5%
voltage distribution line, we deemed that result significant. We found that the comparison of
spontaneous dMSN and iMSN STAs yielded no significant difference. However, the evoked
STA traces between dMSNs and iMSNs were markedly different (Fig. 4F). Furthermore, the
additional input from the sensory cortex seems to have specifically targeted dMSNs, as their
STA traces varied significantly between the spontaneous and evoked conditions, whereas no
major change was observable for the same comparison of iMSNs (Fig. 4G).
Thus, the results of our STA analysis also support the notion that dMSNs receive stronger
11
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
Fig 3. No difference in effective membrane time constant between dMSNs and iMSNs in up-states. AExample of the
effective membrane time constant estimation in a dMSN, for all up-states with the mean membrane potential falling into a single
0.5 mV voltage bin. PSD estimates of individual up-states (grey) were averaged and smoothed (black trace), and the initial
membrane time constant
τini
m
was estimated as the point where the maximal power decreased by
3 dB
(black dashed line).
B
Two-dimensional representation of the matrix used in the τmcorrection procedure. Depending on the average duration of all
up-states within one voltage level, the initial τini
mwas corrected by the appropriate value to obtain the final τmestimate (see
Methods). The black dashed line and the marker represent the data depicted in A.
C
Up-state vs. down-state
τm
for all neurons
regardless of the cell type or physiological condition; the dashed line represents equality. It is clear that τmin the up-states is
smaller than in the down-states, indicating a high-conductance regime due to synaptic bombardment, similar to that in
neocortical neurons (Paré et al., 1998; Destexhe et al., 1999; Léger et al., 2005).
D
There is no significant difference in up-state
τm
between dMSNs and iMSNs, either in control or 6OHDA conditions. This suggests that the differences in up-state membrane
power are not the result of differences in membrane dynamics between dMSNs and iMSNs. In down-states, 6OHDA dMSNs had
higher τmthan the control cells (p= 0.048). Data are shown as mean±SEM. Control dMSNs and iMSNs are in red and blue,
respectively, whereas 6OHDA dMSNs and iMSNs are in light red and light blue, respectively. *p < 0.05
synaptic input than iMSNs in healthy animals.
Discussion
Here we provided evidence that in vivo dMSNs receive stronger synaptic input than iMSNs and
that this difference is attenuated in dopamine-depleted animals. These findings were based on
two observations: (1) dMSNs showed significantly higher spectral power than iMSNs in the
up-states, especially in the higher-frequency bands (Fig. 2C,D), and (2) in both spontaneous
and stimulus-induced spikes dMSNs membranes depolarized faster than iMSNs before reaching
12
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
13
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
spike-threshold, as revealed by their STAs (Fig. 4). These results provide support for the
theoretical prediction that direct-pathway MSNs in healthy state animals receive stronger
synaptic input than iMSNs (Bahuguna et al., 2015). In addition, we showed by spectral
analysis that the effective membrane time constant of MSNs during up-states is significantly
shorter than in down-states, indicating that synaptic inputs affect the membrane conductance
to a larger extent than hyperpolarization-activated conductances mediated by the Kir channels
in MSNs.
Paired recordings in slices revealed that iMSNs form more and stronger synaptic connections
onto dMSNs than they receive from them (Taverna et al., 2008; Planert et al., 2010). In
addition, fast spiking interneurons also form more connections onto dMSNs than onto iMSNs
(Planert et al., 2010; Gittis et al., 2010). Given these differences in connectivity, Bahuguna
et al. (2015) predicted that dMSNs must receive more or stronger excitatory inputs if both
dMSNs and iMSNs were to be co-activated (Cui et al., 2013) or have comparable activity
levels in both ongoing and stimulus evoked activity (Sippy et al., 2015), as has been observed
experimentally. Consistent with this prediction, Parker et al. (2016) showed that, in healthy
animals in vitro, dMSNs receive stronger excitatory input from both thalamo-striatal and
cortico-striatal projections. They also showed that in dopamine-depleted mice thalamo-striatal
inputs to iMSNs become stronger than to dMSNs, whereas cortico-striatal projections remain
largely unchanged.
Here, we show that the disparity between the excitatory inputs to dMSNs and iMSNs is also
maintained in vivo in the up-states, which closely resemble the awake state of an animal
(Destexhe et al., 2003; Haider et al., 2012). In our analysis we assumed that larger fluctuations
in the membrane potential are a reflection of stronger synaptic weights and/or correlated inputs.
Because both thalamus and cortex are co-activated in the up-states, we cannot distinguish
between thalamo-striatal and cortico-striatal inputs. However, by selectively silencing thalamic
inputs to the striatum (using optogenetic or chemogenetic approaches) it should be possible
to determine the relative contributions of thalamo-striatal and cortico-striatal inputs in vivo
following our approach. Another major limitation of our analysis is that we cannot separate
excitatory from inhibitory inputs. In fact, an increase in either type of synaptic inputs can
increase the membrane potential variance (Kuhn et al., 2004). However, our comparison of
STAs suggests that dMSNs are more likely to receive stronger excitatory inputs because during
both, spontaneous and stimulus-evoked activity, dMSNs depolarize faster to the action-potential
threshold than iMSNs (Fig. 4).
Fig 4 (preceding page). dMSNs accelerate faster towards firing threshold than iMSNs when receiving input from barrel
cortex. AExample of the estimation of the firing threshold and extraction of the pre-spike voltage trace. Top: membrane
potential of a dMSN in the up-state; bottom: its first-order derivative dV/dt. The spiking threshold was determined as the
highest voltage deflection seen in the derivative that didn’t produce a spike (black dashed line); when the derivative crossed the
threshold, this marked the start of an action potential (red dot). The voltage trace during 12 ms preceding the spike onset is
marked in green. BExpanded view of the shaded area in A.CExample of calculating the spike-triggered average (STA, red
curve) for the neuron in A. Gray traces are
12 ms
pre-spike intervals from individual up-states producing a spike.
D
Comparison
of grand-average STAs of dMSNs and iMSNs in control conditions (thick lines in red and blue, respectively) when action
potentials were generated by spontaneous activity. Faint red and blue traces show STAs for individual neurons of corresponding
MSN types. All traces were aligned to spike onset. Error bars represent SEM. ESame as in D, but the action potentials were
generated by whisker stimulation and synaptic input from the barrel cortex. Note that the grand-average STA of dMSNs is
accelerating faster toward spike onset, indicating stronger synaptic input to these neurons. FThe permutation test shows that
evoked dMSN and iMSN STAs differed significantly, by falling in the bottom and top
2.5%
of voltage distributions, respectively.
However, no such difference was observed for spontaneous traces.
G
While there was no marked difference between spontaneous
and evoked iMSN STAs, dMSN traces differed significantly across the two conditions.
14
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
The size of membrane potential fluctuations is affected by the mean membrane potential which
determines the synaptic driving force and the membrane time constant (Kuhn et al., 2004).
In our data we did not find a significant difference between the mean up-state membrane
potentials of dMSNs and iMSNs in either control or 6OHDA conditions (in all cases
p >
0
.
138,
data not shown), and no significant difference in up-state
τm
estimations either (Fig. 3D). The
latter is contrary to previous findings in vitro, where measurements of whole-cell capacitances
and input resistances of dMSNs and iMSNs suggest that their membrane time constants are
different (Gertler et al., 2008; Fieblinger et al., 2014). It should be born in mind, however,
that under in vivo conditions the membrane properties are affected by the ongoing synaptic
activity (Kuhn et al., 2004) and that synaptic inputs can easily overcome the differences in
the neuron membrane properties measured in vitro (Destexhe et al., 2003, 2007).
Furthermore, our data shows that in MSNs the effective membrane time constant in the
up-states is on average
46 %
smaller than in the down-states (Table 1, Fig. 3C), indicating a
high-conductance state in striatal neurons in the presence of synaptic inputs, similar to that of
neocortical neurons (Destexhe et al., 2003; Léger et al., 2005; Destexhe et al., 2007). That is,
in the up-states, the membrane time constant is strongly influenced by synaptic inputs.
Finally, our results also provide new insights into how dopamine affects the excitatory inputs
to the MSNs. We found that in dopamine-depleted animals both dMSNs and iMSNs receive
similar amounts of excitatory inputs. These results are consistent with our previous findings
that in healthy animals dMSNs exhibit stronger response to contralateral sensory stimulation
than iMSNs, and that these differences are diminished in dopamine-depleted mice (Reig and
Silberberg, 2014; Ketzef et al., 2017). Thus, our results and previous findings (Parker et al.,
2016; Ketzef et al., 2017) suggest that dopamine is important to maintain the difference in
the excitatory inputs to dMSNs and iMSNs. In the absence of dopamine, excitatory inputs to
dMSNs are weaker and the indirect pathway is active for a much larger range of cortico-thalamic
inputs, resulting in dysfunctional action selection and action initiation (Bahuguna et al., 2015).
While we have provided evidence for stronger total synaptic input to the dMSNs as compared
to iMSNs, it is still unclear whether the extra excitation to the dMSNs is due to stronger
cortical and/or thalamic inputs. Furthermore, it also not clear whether the larger membrane
potential fluctuations in dMSNs are due to stronger synapses or to higher input correlations.
More dedicated experiments involving selective correlated activation of cortical and thalamic
neurons will help resolving these questions.
Acknowledgements
We thank Dr. Jyotika Bahuguna, Dr. Rita Almeida, Matthijs Dorst, Michael Zohar and
Yvonne Johansson for their technical support and advice. This work was funded in parts by:
the EU Erasmus Mundus Joint Doctorate Program ’EUROSPIN’, The International Graduate
Academy (IGA) of the Freiburg Research Services (to Marko Filipović), Swedish Research
Council (Research Project Grant, StratNeuro), Parkinsonfonden grant (to Arvind Kumar),
Swedish Research Council (Research Project Grant) (to Gilad Silberberg), the German Research
Foundation (DFG#1086 BrainLinks BrainTools) (to Ad Aertsen and Arvind Kumar).
15
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
References
Bahuguna J, Aertsen A, Kumar A (2015) Existence and Control of Go/No-Go Decision
Transition Threshold in the Striatum. PLOS Computational Biology 11:e1004233.
Cui G, Jun SB, Jin X, Pham MD, Vogel SS, Lovinger DM, Costa RM (2013) Concurrent acti-
vation of striatal direct and indirect pathways during action initiation. Nature 494:
238–242
.
Deng Y, Lanciego J, Goff LKL, Coulon P, Salin P, Kachidian P, Lei W, Del Mar N, Reiner
A (2015) Differential organization of cortical inputs to striatal projection neurons of the
matrix compartment in rats. Frontiers in systems neuroscience 9:51.
Destexhe A, Hughes SW, Rudolph M, Crunelli V (2007) Are corticothalamic ’up’ states
fragments of wakefulness? Trends in Neurosciences 30:334–342.
Destexhe A, Paré D, Willmore BDB, Cooke JE, King AJ, Physiol J (1999) Impact of Network
Activity on the Integrative Properties of Neocortical Pyramidal Neurons In Vivo. Journal of
Neurophysiology 81:1531–1547.
Destexhe A, Rudolph M, Paré D (2003) The high-conductance state of neocortical neurons in
vivo. Nature Reviews Neuroscience 4:1019–1019.
Doig NM, Moss J, Bolam JP (2010) Cortical and Thalamic Innervation of Direct and
Indirect Pathway Medium-Sized Spiny Neurons in Mouse Striatum. Journal of Neuro-
science 30:14610–14618.
Fieblinger T, Graves SM, Sebel LE, Alcacer C, Plotkin JL, Gertler TS, Chan CS, Heiman M,
Greengard P, Cenci MA, Surmeier DJ (2014) Cell type-specific plasticity of striatal projection
neurons in parkinsonism and L-DOPA-induced dyskinesia. Nature Communications 5:
5316
.
Fucke T, Suchanek D, Nawrot MP, Seamari Y, Heck DH, Aertsen A, Boucsein C (2011)
Stereotypical spatiotemporal activity patterns during slow-wave activity in the neocortex.
Journal of Neurophysiology 106:3035–3044.
Gertler TS, Chan CS, Surmeier DJ (2008) Dichotomous Anatomical Properties of Adult
Striatal Medium Spiny Neurons. Journal of Neuroscience 28:10814–10824.
Gittis AH, Nelson AB, Thwin MT, Palop JJ, Kreitzer AC (2010) Distinct Roles of GABAer-
gic Interneurons in the Regulation of Striatal Output Pathways. Journal of Neuro-
science 30:2223–2234.
Haider B, Häusser M, Carandini M (2012) Inhibition dominates sensory responses in the awake
cortex. Nature 493:97–100.
Holm S (1979) A simple sequentially rejective multiple test procedure. Scandianvian Journal
of Statistics 6:65–70.
Katz Y, Yizhar O, Staiger J, Lampl I (2013) Optopatcher-An electrode holder for simultaneous
intracellular patch-clamp recording and optical manipulation. Journal of Neuroscience
Methods 214:113–117.
16
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
Ketzef M, Spigolon G, Johansson Y, Bonito-Oliva A, Fisone G, Silberberg G (2017) Dopamine
Depletion Impairs Bilateral Sensory Processing in the Striatum in a Pathway-Dependent
Manner. Neuron 94:855–865.e5.
Kress GJ, Yamawaki N, Wokosin DL, Wickersham IR, Shepherd GMG, Surmeier DJ (2013) Con-
vergent cortical innervation of striatal projection neurons. Nature Neuroscience 16:
665–667
.
Kuhn A, Aertsen A, Rotter S (2004) Neuronal Integration of Synaptic Input in the Fluctuation-
Driven Regime. Journal of Neuroscience 24:2345–2356.
Léger J, Stern E, Aertsen A, Heck D (2005) Synaptic integration in rat frontal cortex shaped
by network activity. Journal of Neurophysiology 93:281–293.
Lei W, Jiao Y, Del Mar N, A R (2004) Evidence for Differential Cortical Input to Direct
Pathway versus Indirect Pathway Striatal Projection Neurons in Rats. Journal of Neuro-
science 24:8289–8299.
McGeorge AJ, Faull RLM (1989) The organization of the projection from the cerebral cortex
to the striatum in the rat. Neuroscience 29:503–537.
Mukovski M, Chauvette S, Timofeev I, Volgushev M (2007) Detection of active and silent
states in neocortical neurons from the field potential signal during slow-wave sleep. Cerebral
Cortex 17:400–414.
Nisenbaum ES, Wilson CJ (1995) Potassium currents responsible for inward and outward
rectification in rat neostriatal spiny projection neurons. The Journal of neuroscience : the
official journal of the Society for Neuroscience 15:4449–4463.
Papoulis A, Pillai SU (2002) Probability, Random Variables and Stochastic Processes Tata
McGraw-Hill Education, 4th edition.
Paré D, Shink E, Gaudreau H, Destexhe A, Lang EJ (1998) Impact of spontaneous synaptic
activity on the resting properties of cat neocortical pyramidal neurons In vivo. Journal of
neurophysiology 79:1450–1460.
Parker PRL, Lalive AL, Kreitzer AC (2016) Pathway-Specific Remodeling of Thalamostriatal
Synapses in Parkinsonian Mice. Neuron 89:734–740.
Paxinos G, Franklin KB (2004) The mouse brain in stereotaxic coordinates Gulf professional
publishing.
Planert H, Szydlowski SN, Hjorth JJJ, Grillner S, Silberberg G (2010) Dynamics of Synaptic
Transmission between Fast-Spiking Interneurons and Striatal Projection Neurons of the
Direct and Indirect Pathways. Journal of Neuroscience 30:3499–3507.
Reig R, Silberberg G (2014) Multisensory Integration in the Mouse Striatum. Neu-
ron 83:1200–1212.
Santini E, Valjent E, Usiello A, Carta M, Borgkvist A, Girault JA, Herve D, Greengard P, Fisone
G (2007) Critical Involvement of cAMP/DARPP-32 and Extracellular Signal-Regulated Pro-
tein Kinase Signaling in L-DOPA-Induced Dyskinesia. Journal of Neuroscience 27:
6995–7005
.
17
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
Filipović et al. Comparison of total input to striatal MSNs in vivo
Sippy T, Lapray D, Crochet S, Petersen CCH (2015) Cell-Type-Specific Sensorimotor Processing
in Striatal Projection Neurons during Goal-Directed Behavior. Neuron 88:298–305.
Stern Ea, Kincaid AE, Wilson CJ (1997) Spontaneous subthreshold membrane potential
fluctuations and action potential variability of rat corticostriatal and striatal neurons in vivo.
Journal of neurophysiology 77:1697–1715.
Taverna S, Ilijic E, Surmeier DJ (2008) Recurrent Collateral Connections of Striatal
Medium Spiny Neurons Are Disrupted in Models of Parkinson’s Disease. Journal of Neuro-
science 28:5504–5512.
Waters J, Helmchen F (2006) Background Synaptic Activity Is Sparse in Neocortex. Journal
of Neuroscience 26:8267–8277.
Wilson CJ, Kawaguchi Y (1996) The origins of two-state spontaneous membrane potential
fluctuations of neostriatal spiny neurons. The Journal of neuroscience : the official journal
of the Society for Neuroscience 16:2397–410.
18
.CC-BY-NC 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/651711doi: bioRxiv preprint first posted online May. 27, 2019;
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Parkinson’s disease (PD) is a movement disorder caused by the loss of dopaminergic innervation, particularly to the striatum. PD patients often exhibit sensory impairments, yet the underlying network mechanisms are unknown. Here we examined how dopamine (DA) depletion affects sensory processing in the mouse striatum. We used the optopatcher for online identification of direct and indirect pathway projection neurons (MSNs) during in vivo whole-cell recordings. In control mice, MSNs encoded the laterality of sensory inputs with larger and earlier responses to contralateral than ipsilateral whisker deflection. This laterality coding was lost in DA-depleted mice due to adaptive changes in the intrinsic and synaptic properties, mainly, of direct pathway MSNs. L-DOPA treatment restored laterality coding by increasing the separation between ipsilateral and contralateral responses. Our results show that DA depletion impairs bilateral tactile acuity in a pathway-dependent manner, thus providing unexpected insights into the network mechanisms underlying sensory deficits in PD.
Article
Full-text available
Goal-directed sensorimotor transformation drives important aspects of mammalian behavior. The striatum is thought to play a key role in reward-based learning and action selection, receiving glutamatergic sensorimotor signals and dopaminergic reward signals. Here, we obtain whole-cell membrane potential recordings from the dorsolateral striatum of mice trained to lick a reward spout after a whisker deflection. Striatal projection neurons showed strong task-related modulation, with more depolarization and action potential firing on hit trials compared to misses. Direct pathway striatonigral neurons, but not indirect pathway striatopallidal neurons, exhibited a prominent early sensory response. Optogenetic stimulation of direct pathway striatonigral neurons, but not indirect pathway striatopallidal neurons, readily substituted for whisker stimulation evoking a licking response. Our data are consistent with direct pathway striatonigral neurons contributing a “go” signal for goal-directed sensorimotor transformation leading to action initiation. Video Abstract
Article
Full-text available
A typical Go/No-Go decision is suggested to be implemented in the brain via the activation of the direct or indirect pathway in the basal ganglia. Medium spiny neurons (MSNs) in the striatum, receiving input from cortex and projecting to the direct and indirect pathways express D1 and D2 type dopamine receptors, respectively. Recently, it has become clear that the two types of MSNs markedly differ in their mutual and recurrent connectivities as well as feedforward inhibition from FSIs. Therefore, to understand striatal function in action selection, it is of key importance to identify the role of the distinct connectivities within and between the two types of MSNs on the balance of their activity. Here, we used both a reduced firing rate model and numerical simulations of a spiking network model of the striatum to analyze the dynamic balance of spiking activities in D1 and D2 MSNs. We show that the asymmetric connectivity of the two types of MSNs renders the striatum into a threshold device, indicating the state of cortical input rates and correlations by the relative activity rates of D1 and D2 MSNs. Next, we describe how this striatal threshold can be effectively modulated by the activity of fast spiking interneurons, by the dopamine level, and by the activity of the GPe via pallidostriatal backprojections. We show that multiple mechanisms exist in the basal ganglia for biasing striatal output in favour of either the `Go' or the `No-Go' pathway. This new understanding of striatal network dynamics provides novel insights into the putative role of the striatum in various behavioral deficits in patients with Parkinson's disease, including increased reaction times, L-Dopa-induced dyskinesia, and deep brain stimulation-induced impulsivity.
Article
Full-text available
In prior studies, we described the differential organization of corticostriatal and thalamostriatal inputs to the spines of direct pathway (dSPNs) and indirect pathway striatal projection neurons (iSPNs) of the matrix compartment. In the present electron microscopic (EM) analysis, we have refined understanding of the relative amounts of cortical axospinous versus axodendritic input to the two types of SPNs. Of note, we found that individual dSPNs receive about twice as many axospinous synaptic terminals from IT-type (intratelencephalically projecting) cortical neurons as they do from PT-type (pyramidal tract projecting) cortical neurons. We also found that PT-type axospinous synaptic terminals were about 1.5 times as common on individual iSPNs as IT-type axospinous synaptic terminals. Overall, a higher percentage of IT-type terminals contacted dSPN than iSPN spines, while a higher percentage of PT-type terminals contacted iSPN than dSPN spines. Notably, IT-type axospinous synaptic terminals were significantly larger on iSPN spines than on dSPN spines. By contrast to axospinous input, the axodendritic PT-type input to dSPNs was more substantial than that to iSPNs, and the axodendritic IT-type input appeared to be meager and comparable for both SPN types. The prominent axodendritic PT-type input to dSPNs may accentuate their PT-type responsiveness, and the large size of axospinous IT-type terminals on iSPNs may accentuate their IT-type responsiveness. Using transneuronal labeling with rabies virus to selectively label the cortical neurons with direct input to the dSPNs projecting to the substantia nigra pars reticulata, we found that the input predominantly arose from neurons in the upper layers of motor cortices, in which IT-type perikarya predominate. The differential cortical input to SPNs is likely to play key roles in motor control and motor learning.
Article
Full-text available
The basal ganglia are involved in sensorimotor functions and action selection, both of which require the integration of sensory information. In order to determine how such sensory inputs are integrated, we obtained whole-cell recordings in mouse dorsal striatum during presentation of tactile and visual stimuli. All recorded neurons responded to bilateral whisker stimulation, and a subpopulation also responded to visual stimulation. Neurons responding to both visual and tactile stimuli were located in dorsomedial striatum, whereas those responding only to whisker deflections were located dorsolaterally. Responses were mediated by overlapping excitation and inhibition, with excitation onset preceding that of inhibition by several milliseconds. Responses differed according to the type of neuron, with direct pathway MSNs having larger responses and longer latencies between ipsilateral and contralateral responses than indirect pathway MSNs. Our results suggest that striatum acts as a sensory “hub” with specialized functional roles for the different neuron types.
Article
Full-text available
Anatomical studies have led to the assertion that intratelencephalic and pyramidal tract cortical neurons innervate different striatal projection neurons. To test this hypothesis, we measured the responses of mouse striatal neurons to optogenetic activation of intratelencephalic and pyramidal tract axons. Contrary to expectation, direct and indirect pathway striatal spiny projection neurons responded to both intratelencephalic and pyramidal tract activation, arguing that these cortical networks innervate both striatal projection neurons.
Article
The striatum is widely viewed as the fulcrum of pathophysiology in Parkinson’s disease (PD) and L-DOPA-induced dyskinesia (LID). In these disease states, the balance in activity of striatal direct pathway spiny projection neurons (dSPNs) and indirect pathway spiny projection neurons (iSPNs) is disrupted, leading to aberrant action selection. However, it is unclear whether countervailing mechanisms are engaged in these states. Here we report that iSPN intrinsic excitability and excitatory corticostriatal synaptic connectivity were lower in PD models than normal; L-DOPA treatment restored these properties. Conversely, dSPN intrinsic excitability was elevated in tissue from PD models and suppressed in LID models. Although the synaptic connectivity of dSPNs did not change in PD models, it fell with L-DOPA treatment. In neither case, however, was the strength of corticostriatal connections globally scaled. Thus, SPNs manifested homeostatic adaptations in intrinsic excitability and in the number but not strength of excitatory corticostriatal synapses.
Article
Movement suppression in Parkinson’s disease (PD) is thought to arise from increased efficacy of the indirect pathway basal ganglia circuit, relative to the direct pathway. However, the underlying pathophysiological mechanisms remain elusive. To examine whether changes in the strength of synaptic inputs to these circuits contribute to this imbalance, we obtained paired whole-cell recordings from striatal direct- and indirect-pathway medium spiny neurons (dMSNs and iMSNs) and optically stimulated inputs from sensorimotor cortex or intralaminar thalamus in brain slices from control and dopamine-depleted mice. We found that dopamine depletion selectively decreased synaptic strength at thalamic inputs to dMSNs, suggesting that thalamus drives asymmetric activation of basal ganglia circuitry underlying parkinsonian motor impairments. Consistent with this hypothesis, in vivo chemogenetic and optogenetic inhibition of thalamostriatal terminals reversed motor deficits in dopamine-depleted mice. These results implicate thalamostriatal projections in the pathophysiology of PD and support interventions targeting thalamus as a potential therapeutic strategy.
Article
Optogenetics has rapidly become a standard method in neuroscience research. Although significant progress has been made in the development of molecular tools, refined techniques for combined light delivery and recording in vivo are still lacking. For example, simultaneous intracellular recording and light stimulation have only been possible by using two separate positioning systems. To overcome this limitation, we have developed a glass pipette holder which contains an additional port for the insertion of an optical fiber into the pipette. This device, which we called "optopatcher" allows whole cell patch-clamp recording simultaneously with direct projection of light from the recording pipette. The holder spares the use of an additional manipulator and, importantly, enables accurate, stable and reproducible illumination. In addition, replacement of standard pipettes is done as easily as with the available commercial holders. Here we used the optopatcher in vivo to record the membrane potential of neurons from different cortical layers in the motor cortex of transgenic mice expressing channelrhodopsin-2 under the Thy1 promoter. We demonstrate the utility of the optopatcher by recording LFP and intracellular responses to light stimulation.